Pediatric Endosurgery & Innovative Techniques Volume 7, Number 2, 2003 Mary Ann Liebert, Inc. Laparoscopic or Endoscopic Gastrostomy in Children: Comparison of Two Methods H. STEYAERT, L. CARFAGNA, M.A. LEMBO, E. TREVINO, and J.S. VALLA ABSTRACT Background and Purpose: Gastrostomy tubes are routinely placed endoscopically. However, complications are being reported with increasing frequency. To assess the relative efficacy of endoscopic and laparoscopic placement, we prospectively compared two groups of children in whom a gastrostomy tube was placed by either technique. Patients and Methods: Thirty-three gastrostomies were successively placed in 29 children. For patients with gastroesophageal reflux, an antireflux procedure was combined with a laparoscopic gastrostomy (lapg) placement (14 children). In the other cases, a percutaneous endoscopic gastrostomy (PEG) was performed (19 children). Results: The overall complication rate was 30% (37% for PEG and 21.5% for lapg). Major complications in the PEG group included fistula, irreducible restenosis, and peritonitis after button placement. Major complications in the lapg group consisted mainly of parietal infections. Conclusions: Gastrostomy tube placement is a serious operation that is associated with a high rate of complications whatever technique is used. Our series suggests that lapg is the best way to reduce morbidity. Better visualization makes it possible to choose the position of the gastrostomy in the stomach and abdominal wall accurately, avoid perforation of the colon or an intestinal loop, and fix the stomach to the rectal sheath securely. To decrease the incidence of parietal infections in lapg, we propose (1) the use of prophylactic antibiotics and (2) enlargement of the trocar opening. Primary button placement also decreases the rate of parietal complications. INTRODUCTION GASTROSTOMY IS WIDELY USED IN PEDIATRICS to establish supplemental feeding in children who fail to thrive or who have swallowing disorders, especially in association with neurologic impairment. Percutaneous endoscopic gastrostomy (PEG) has for years been considered the minimally invasive technique of choice in such cases. However, recent studies have reported a high morbidity rate and the development of gastroesophageal reflux in up to 30% of patients. 1 Complications range from minor wound infections to peritonitis (even after secondary button insertion) or digestive fistula. Pediatric Surgery, Fondation Lenval pour Enfants, Nice, France. Presented at the 10 th Annual Congress for Endosurgery in Children, Brisbane, Australia, March, 2001. 141
STEYAERT ET AL. The aim of this study was to compare PEG with laparoscopic gastrostomy (lapg) in a prospective manner and to identify risk factors for the development of complications after each method. MATERIALS AND METHODS All children who underwent gastrostomy placement at our institution after January 1996 were followed prospectively. All were evaluated preoperatively for gastroesophageal reflux by 24-hour ph monitoring. The patients who required antireflux surgery as determined by abnormal results of 24-hour ph monitoring underwent lapg placement at the end of the operation, and the others underwent PEG placement. Thirtythree gastrostomies (19 PEGs and 14 lapgs) were placed in 29 children. The age of the patients ranged from 1 month to 18 years (average, 5 years). In cases of PEG placement, the size of the gastrostomy tube depended on the age of the patient. In all PEG cases, we used the Mic-Key system (Ballard Medical, Draper, Utah). The Mic-Key button was placed secondarily after 2 or 3 months. In cases of lapg placement, a Nutricia tube with a balloon (Nutricia Clinical Care, Trowbridge, Wiltshire, United Kingdom) was placed primarily; a Mic-Key button was placed secondarily after the same interval as in PEG. The last three lapg cases underwent direct button placement. The indications for PEG and lapg are listed Table 1. A prolonged hospital stay as a consequence of the use of the technique and a repeated hospitalization or procedure that required the administration of general anesthesia were considered to be major complications. Minor complications were not considered in this study. Operative technique In the PEG cases, we used the method described by Khattak et al. 1 All the gastrostomy procedures were supervised by a pediatric gastroenterologist and a pediatric surgeon and were performed with the patient under general anesthesia. In the lapg cases, the stomach was grasped at the ventral border between the corpus and antrum with a 3-mm forceps after a trocar had been placed under optical control at the end of the antireflux operation. The stomach was fixed to the internal rectal sheath, and a tube was inserted into the stomach and secured with a purse-string suture. RESULTS Follow-up (1 5 years; average, 2.5 years) information was available for all patients. The overall complication rate was 30% (37% for PEG and 21.5% for lapg). Major complications Stomal infection. Two major stomal infections occurred in the lapg group and one in the PEG group. In two of them (one in each group), the tube eventually had to be removed, and another tube was placed af- TABLE 1. INDICATIONS FOR GASTROSTOMY TUBE PLACEMENT PEG LapG Neurologic impairment 15 12 Esophageal atresia 2 KNO malformation 2 Cerebral tumor 1 Cystic fibrosis 1 LapG, laparoscopic gastrostomy placement; PEG, percutaneous endoscopic gastrostomy. 142
ter complete healing. In the case in the PEG group, after a new PEG had been placed a few centimeters away, a fistula formed between the two openings that was finally closed laparoscopically. In the third case, tube removal was not required, but an infection developed that necessitated a prolonged hospitalization of more than 2 months. Peritonitis. In two patients in whom button placement had been uncomplicated, peritonitis developed 2 1/2 and 3 months after PEG. The two patients received alimentation just after button placement and rapidly became symptomatic. Laparoscopy was undertaken the same day to perform thorough lavage and drainage. In both cases, the stomach adhered to the peritoneal wall for only half of the circumference, and the button was placed directly into the abdominal cavity. The stomach wall around the button was sutured to the anterior wall by laparoscopy. Recovery was uneventful in the two cases. In another case, acute peritonitis developed following a button change (the second button change for this patient) more than 1 year after PEG. Fistula, intestinal perforation. One patient presented with fever and signs of peritonitis 5 days after PEG. At laparotomy, a gastrocolocutaneous fistula was found. After colonic repair, a Stamm gastrostomy was placed. In a baby who underwent lapg, the tube was primarily inserted into an intestinal loop. A contrast study performed at the end of the operation revealed the incorrect placement. The intestinal loop was repaired, and the tube was inserted into the stomach. The recovery was uneventful in both cases. Failure to conduct the PEG procedure. In one case, the position of the stomach made transillumination impossible, and the PEG procedure was abandoned. A lapg was successfully performed a few days later; this made it necessary to administer general anesthesia for a second time to a patient with scoliosis. Stomal stenosis. The tube of one PEG patient was lost at home. The patient was not brought to the hospital until 3 days later, and stenosis had developed. A new PEG placement was required because the stenosis could not be corrected. Minor complications LAPAROSCOPIC OR ENDOSCOPIC GASTROSTOMY IN CHILDREN Minor complications consisted mainly of peristomal infections or bleeding from the margin of the stoma. Local treatment was sufficient in most cases. No perioperative mortality occurred. DISCUSSION The endoscopic placement of feeding tubes is now accepted as the preferred technique for establishing long-term enteral feeding. However, complications are being reported with increasing frequency. 1 Furthermore, children who undergo PEG require general anesthesia, and the benefits of PEG in comparison with lapg are not proven. 2 Of course, our study is not a direct comparison. However, antireflux surgery never interfered with button placement, although the longer operative time may increase the infection rate. Our series suggests that lapg is associated with fewer major complications. Visualization of the stomach and adjacent organs makes it possible to choose the position of the gastrostomy accurately and avoid the most important complication of PEG, which is gastrocolocutaneous or other types of fistula. 3 4 The location of the gastrostomy tube appears to be important; lesser curvature gastrostomy is reported to be associated with less reflux than greater curvature gastrostomy. 2 Gastroesophageal reflux developed in two (10.5%) of our patients after PEG. Both required a Nissen procedure a few month later. In such cases, the presence of a gastrostomy makes the operation more difficult, but removal of the button was never necessary. However, the operative time was increased. The development of gastroesophageal reflux after PEG is well-known. Most authors estimate that 10% to 20% of patients require antireflux surgery after PEG. 1 5 One of the ways to address this problem is to determine whether nasogastric tube feeding is well tolerated or not. In cases of good tolerance, PEG or lapg without an antireflux procedure is recommended. 5 143
Perfect fixation of the stomach to the gastric wall prevents the development of peritonitis when the secondary button is placed. Dislodgement of the gastrostomy after PEG is another major complication. 6 Intraperitoneal spillage leads to peritonitis and can be fatal. We never encountered this type of complication after primary lapg button placement. Only lapg allows a perfect view of the stomach wall. In cases of button change after PEG, we now recommend systematic opacification of the stomach through the button before realimentation is undertaken. Migration of the PEG (or parts of it) has also been described. We had one case without problems because no fistula developed. However, intestinal fistula has been reported. 4 The use of a larger gauge may prevent this. 7 Fixation to the wall by lapg is probably more secure and so avoids this type of complication. If all the gastrostomies in our series had been placed laparoscopically, most of the significant complications we encountered would have been avoided. The most important complication of lapg is parietal infection. To decrease the rate of major parietal infections, we propose enlarging the trocar opening after lapg to permit correct suture placement and avoid the local decrease in the blood supply to the stomach wall that results from compression. 1 The use of prophylactic antibiotics at the time of gastrostomy tube placement also decreases the infectious complication rate. Antibiotics were not used in our series (PEG or lapg). Moreover, in our last three cases of lapg placement, a button gastrostomy was created primarily. The operation was safe, and most of the secondary complications that appeared at the time of conversion from tube gastrostomy to button gastrostomy resolved without treatment. Because laparoscopy is known to cause fewer adhesions than laparotomy, this argument may no longer be used to promote PEG. However, PEG would probably be chosen for patients who had previously undergone abdominal surgery. With lapg, other procedures can be performed simultaneously for example, antireflux surgery or pyloroplasty. 2 The preventive treatment of gastroesophageal reflux during anesthesia for gastrostomy is debatable. We generally recommend a Nissen procedure at the time of gastrostomy in neurologically impaired patients. They usually have gastroesophageal reflux that is frequently associated with esophagitis. However, pediatricians are not always convinced. In this series, we recommended 24-hour ph monitoring for each child who was evaluated for gastrostomy. 8 Antireflux surgery was always recommended in cases with abnormal results of ph monitoring. The technique of lapg is also debatable. 3,9,10 Early in our experience, we performed both gastrostomy and laparoscopic fundoplication through the same trocar sites. The location of the trocars was not always ideal for exteriorization of the stomach. Eventually, we inserted another trocar at the correct site, and the complication rate decreased immediately. For this reason, we now think, as do others, that a two-trocar method with enlargement of the trocar site after exteriorization of the stomach is the technique of choice for gastrostomy tube placement. 11 In cases of gastroesophageal reflux surgery, we advocate placement of a new trocar if the stomach is difficult to grasp and bring to the anterior wall through the operating trocars. Experience has also shown that primary button placement greatly decreases the need for wound care and the mechanical problems associated with tubes. 12 STEYAERT ET AL. CONCLUSION Gauderer et al. 13 revolutionized gastrostomy tube placement with the introduction of PEG. Since then, several reports have described complications caused mainly by (1) a partly blind technique and (2) failure to fix the stomach to the abdominal wall. Laparoscopy is a good minimally invasive method for dealing with these two problems. However, we must never forget that gastrostomy placement is associated with significant morbidity whatever technique is used. 1 14 Perfect technique and a long postoperative follow-up are mandatory. Laparoscopic gastrostomy with a two-trocar technique and primary insertion of the button is now our method of choice. 12 In each case, we propose preoperative 24-hour ph monitoring to determine the presence of gastroesophageal reflux. If the patient has reflux, we advocate laparoscopic antireflux surgery with lapg at the end of the operation. 144
LAPAROSCOPIC OR ENDOSCOPIC GASTROSTOMY IN CHILDREN Enlargement of the trocar site after exteriorization of the stomach is essential to avoid the parietal infections that were the main complication in our lapg series. REFERENCES 1. Khattak IU, Kimber C, Kiely EM, Spitz L. Percutaneous endoscopic gastrostomy in pediatric practice: complications and outcome. J Pediatr Surg 1998;33:67 72. 2. Georgeson KE. Laparoscopic versus open procedures for long-term enteral access. Nutr Clin Pract 1997;12:S7 S8. 3. Humphrey GME, Najmaldin A. Laparoscopic gastrostomy in children. Pediatr Surg Int 1997;12:S01 S04. 4. Kubiak R, Wilcox DT, Spitz L. Gastrojejunal fistula after insertion of percutaneous endoscopic gastrostomy. J Pediatr Surg 1999;34:1287 1288. 5. Isch JA, Rescorla FJ, Scherer LRT III, West KW, Grosfeld JL. The development of gastroesophageal reflux after percutaneous endoscopic gastrostomy. J Pediatr Surg 1997;32:321 323. 6. Pofahl WE, Ringold F. Management of early dislodgement of percutaneous endoscopic gastrostomy tubes. Surg Laparosc Endosc 1999;4:253 256. 7. Simon T, Fink AS. Recent experience with percutaneous endoscopic gastrostomy/jejunostomy for enteral nutrition. Surg Endosc 2000;14:436 438. 8. Sullivan PB. Gastrostomy feeding in the disabled child: When is an antireflux procedure required? Arch Dis Child 1999;81:463 464. 9. Kellnar ST, Till H, Böhm R. Laparoscopically assisted performance of gastrostomy: A simple, safe and minimally invasive technique. Eur J Pediatr Surg 1999;9:297 298. 10. Stylianos S, Flanigan LM. Primary button gastrostomy: A simplified percutaneous, open, laparoscopy-guided technique. J Pediatr Surg 1995;30:219 220. 11. Ng PC. Laparoscopic gastrostomy: A simple way to feed. Surg Laparosc Endosc 1994;4:463 464. 12. Rothenberg SS, Bealer JF, Chang JHT. Primary laparoscopic placement of gastrostomy buttons for feeding tubes. Surg Endosc 1999;13:995 997. 13. Gauderer MWL, Ponsky JL, Izant RJ. Gastrostomy without laparotomy: A percutaneous endoscopic technique. J Pediatr Surg 1980;15:872 875. 14. Ho HS, Ngo H Gastrostomy for enteral access. A comparison among placement by laparotomy, laparoscopy, and endoscopy. Surg Endosc 1999 ;13:991 994. Address reprint requests to: Dr. H. Steyaert, MD Pediatric Surgery Fondation Lenval 57, Avenue de la Californie F06200, Nice, France E-mail: henri.steyaert@lenval.com 145